Conan Kudo

Why would they have to divest anything? They'd still be smaller than VZW, ATT.

Because combining Sprint's >200MHz of spectrum with T-Mobile's ~80MHz of spectrum is more than twice the spectrum screen limit for any operator. Further subdividing into spectrum classes, the only spectrum they wouldn't be required to divest is 700MHz, ESMR 800MHz, and Cellular 850MHz spectrum. They'd have to get rid of either some PCS or AWS, and a great deal of its BRS+EBS would have to be divested, because there's no rational public interest benefit to Sprint controlling damn near the entire 194MHz of the 2.6GHz band.

Precisely because Sprint has so much BRS/EBS, and will have significantly deployed in their network it before this merger ever closes. It's also already baked in to newer Sprint phones. Also, getting rid of leased EBS will be less helpful with FCC approval than agreeing to give up AWS.

 

To make this work as a business proposition, they are going to have to thin their combined site count. I'm assuming the network gear and UE that does get scrapped or resold will be the TMUS stuff. They don't have to get rid of all TMUS sites, just eliminate ones with overlap and where Sprint has plenty of capacity, post NV. Once you decide that, then the AWS spectrum is the logical spectrum to divest, especially given its resale value. Lots of regionals would love to have some, as well as ATT and VZW.

By that logic, PCS would be a better divestment, because everyone uses PCS. Every operator would love to get their hands on parts of Sprint's PCS D+B+E+F spectrum throughout the country, since it's smack in the middle of the band and goes well with all currently deployed technologies except LTE from an ecosystem perspective.

Thai food

 

I think there is some confusion. When I say that Sprint would divest the AWS spectrum, they won't have to do it immediately. They will be given maybe 2-3 years to reach agreements to sell it (Just like Verizon was given to sell its lower 700Mhz spectrum as a condition of getting CableCo's AWS spectrum). Here is how I would handle integrating TMUS if I were managing Sprint.

 

Almost immediately:

• Sprint stops selling new TMUS phones. All new UE will be Sprint gear. 
• Sprint makes TMUS LTE (AWS) "native" for UE that supports it. I imagine the iPhone 5S or the Nexus 5 could take advantage of this with a simple software update? (AWS is being used as a "bridge" while the merger and the TD-LTE rollout are completed.

As more TMUS subscribers move to Sprint handsets/UE:

• Sprint gradually repurposes TMUS PCS spectrum for Sprint LTE. 
• Sprint strategically 'thins" TMUS sites, converting selected TMUS sites to NV sites.
• Announce a firm shutdown date for the TMUS network. 
• Reach agreements with buyers for AWS spectrum (maybe they get some immediately the remainder when the TMUS network is shut off)

At the end (about three years from the date of the merger):

• All remaining TMUS sites are shut down
• Remaining AWS spectrum is transferred to whomever purchased it.

 

Why in the world would Sprint be so stupid as to give up such bandwidth for a paltry LTE FDD network on PCS? Especially when PCS isn't even the band of choice for LTE anyway!

Given the layout of PCS spectrum, it makes more sense to retain UMTS service on PCS and accelerate the shift of UMTS to PCS and LTE to AWS. It is more likely that Sprint's PCS G block would be divested rather than AWS, because it's an island band that simply isn't very useful.

CDMA to UMTS conversions are quite commonplace and very easy to do with today's gear. No one sells CDMA gear that doesn't have support for UMTS because there have been many situations where CDMA operators have chosen to switch platforms as part of a strategy or a government order.

More importantly, AWS is the global roaming band for the Americas. Removing that would be foolish. AWS has some special characteristics that make it uniquely qualified to be a global band, and every operator in the Americas knows this.

All of that is unnecessary because Sprint owns 100+ Mhz of EBS/BRS on average nationwide, and if they deploy 60-80Mhz they can still divest 20-40Mhz. That's already an insane amount of capacity! Why would you ever "thin out" fully capable T-Mobile's LTE network that Sprint subs can start benefitting immediately? I don't get it.

 

No device can use more than 40MHz due to PA, filter, and antenna restrictions on user equipment. It makes very little sense to have the entire band. BRS+EBS is an easier divestment to make, as well as PCS G block.

Time Magic

World Cup

It's not quite available to everyone on TMo after all. AMR-WB only works when connected to a modernized tower, and not all of their urban HSPA+ footprint is upgraded with modernized NSN or Ericsson equipment.

That's not strictly true. AMR-WB is supported on all UMTS cell sites that T-Mobile has. It's even supported on newer GSM cell sites, as well. The limiter is whether or not your handset supports the codec. Most handsets released in the last two years have the support built-in.

Grand Line

And the final thing is that there are a lot of sites not upgraded within their launch cities. Just the same as Sprint. But they don't get called out for it.

T-Mobile doesn't get called on it largely because it's not that noticeable to its subscribers. The reason it was noticeable with Verizon and Sprint is that the fallback is very painful. AT&T and T-Mobile both offer a superior fallback experience, and they actively work to improve performance on both levels of the network. 

HSPA+21 and HSPA+42 can provide a substantially better experience than EvDO Rev A. If Rev B was deployed, maybe that might be a different story, but EvDO is an dead-end product anyway.

However, there's some uncertainty on whether the CA will be with Band 17+2 or Band 12+2. Both are supported in the spec, but depending on how fast the agreement gets implemented, AT&T may do the latter rather than the former...

That's kinda crazy.

 

Why did they limit themselves to Asia only?

Nortel locked them out of the Americas, and Lucent kept them out of Europe. After both suffered their fates, Samsung was able to come into the global infrastructure business. It also helped that Samsung devices were becoming very popular, too.

No names but a certain guy with three parallel blue dashes on his door told me several times how much loves Samsung equipment over his own and couldn't wait to get back to integrating their equipment again.

Samsung Networks has been in the business since the late 1980s. However, it entered Europe and the Americas fairly recently. It had restricted itself to Asia prior to that...

I still think there's big advantage in using bands 26 and 41 in concert. 

 

But yeah, it is kind of a puzzle for SoftBank to synchronize their band puzzle here and in Japan. I'm not familiar with their band allocation in Japan other than Platinum Band which matches up with 900 MHz European GSM/UMTS if I'm correct. 

 

VoLTE is close enough that it makes sense for Sprint to hang on with CDMA2000 in the interim, which isn't long at all. 

SoftBank's network uses the following technology and bands:

• UMTS 900/1500/1700/2100 (3GPP bands 8, 11, 9, 1)
• LTE 700/1800/2100 (3GPP bands 28, 3, 1)

Most of their handsets support UMTS 900/1500/1900/2100, as UMTS 1700 is being dropped for LTE 1800.

The reason UMTS 1900 (3GPP band 2) is supported on a vast majority of their handsets is because there's a lot of roaming with Latin America and Europe/Japan. Both Europe and Japan have been active in the Latin American economy for quite some time now. PCS was the first band that UMTS was deployed on in the Americas, though now it's on 850 and 1900. It's also for this reason that the vast majority of European handsets are tri-band UMTS 900/1900/2100. It's also fairly easy to support it, too.

But what would be the point?! It would be a tremendous additional expenditure both for its customers and for Sprint. You are already moving to an all LTE network. Why add additional costly intermediate steps? In the areas where Sprint has a full 7x7 of 800 spectrum, which is most of the US outside the SE (SoLinc) and the IBEZ, I suppose you could go from one 5x5 of LTE to a pair of 3x3 LTE channels, You would get a negligible gain of capacity though. I imagine the 1xA channels of CDMA on 800 will be the last part of the CDMA network Sprint shutters. To take advantage of the spectral efficiencies of FD-LTE, you really want to have spectrum allocations that are 5x5, 10x10, 15x15, or 20x20. Since the PCS spectrum was all auctioned as 5x5, 10x10 or 15x15, this fits perfectly. Sprint's 800 and the CLR 850 band allocations aren't nice contiguous 5Mhz paired allocations. Thus you might as well use the leftover slivers to continue to run a few channels of CDMA right up until the bitter end.

Sprint has plenty of non-contiguous allocations that prevent any large efficiency with PCS LTE. The only shot Sprint has with PCS LTE to go up to a larger channel is if it gets PCS H block. But that doesn't matter either. SoftBank does not care about the PCS network (because it is CDMA on PCS A-F, there's no confidence in a band 25 LTE ecosystem right now, and there are no synergies to realize here because of CDMA instead of UMTS on PCS A-F). It doesn't care about the ESMR network. It cares about the IMT-E TDD network. SoftBank has already enforced a strategic change to Sprint that has refocused capacity efforts around Band 41 instead of the other bands.

Sprint has been offering UMTS band 2 service support on a large number of its devices for a few years now, so it could do that. ESMR would likely retain a single CDMA 1X-Advanced carrier alongside the 3MHz FDD LTE carrier. As you said, there's no real reason to not retain the legacy CDMA 1X-A service on ESMR until the bitter end. And it wouldn't cost that much more to support it, since nearly the entire system is already in place. Nearly all of the equipment deployed in Network Vision can be reused for a UMTS/LTE infrastructure on ESMR/PCS/TD-IMT-E spectrum.

True, the iPhone 5S & 5C would work on both networks equally well (except for LTE band 41). From here on out every Sprint phone should include GSM/UMTS radios and perhaps every handset that SoftBank sells should include support for Sprint's CDMA & LTE networks.

 

It's also interesting that until about a year or two ago that Sprint had arguably out of the big four the most compelling international data plan at $40 for unlimited regardless of whether it was 2G or 3G. Then they abandoned that for (much) more expensive options. Hopefully this maneuver from t-mobile will force them to rethink their international offerings a bit.

 

Sent from my SPH-L900 using Tapatalk

 

SoftBank will never change its requirements to include CDMA. Adding support for Sprint's and T-Mobile's LTE networks will come very soon, but there's absolutely no reason to include CDMA on its phones (excluding iPhones). It is a UMTS/LTE operator, not a CDMA one.

I like this discussion. I do see some positive things out of the Eurasian discussion. I believe interoperability is a great thing. Interoperability is how the Bell System became the best phone system in the world. A lot of people seem to forget that.

 

That said, I don't envy their increased regulation and not having a common policy on auctions so the entire continent can clear up 800 spectrum at once. The EU messes up by over-regulating in my point of view. I differ from AJ's view in that I don't point the finger at the GSMA or 3GPP, bodies where the US carriers have a more increasing role as it is anyway. I point that finger at the European Commission.

That is definitely true. I'm glad you brought up the Bell System. For all of its faults, it was a good thing from a technological perspective. The Bell System eventually grew, incorporated features from rival systems, and became the gold standard for a telephone system. The Bell System did it by incorporating the feedback from those in various markets it expanded into (Mexico, Japan, Europe, etc.). That made it a better system, overall. The 3GPP does the same thing today with UMTS and LTE.

As for spectrum release issues in Europe, the EC is working on streamlining the process this time around with the APT700 subset that will be implemented for its second digital dividend spectrum release in about a year. The EC messed up this time with the CEPT800 plan because they allowed the individual countries to have their own timetables on digital TV conversion.

To be fair, part of the problem was the differing TV standards used throughout the region covered by CEPT. Not all of them used PAL (which tended to move toward lower frequency systems anyway). Several used SECAM instead. The countries also argued for years about which system to move to until they EU put its foot down and told them which system to use. These problems won't come up again this time around, so I think it'll move along much more smoothly.

Europe isn't "behind" so much as they're avoid our mistakes.

 

Verizon, Sprint, and AT&T's LTE all being on separate bands that devices can't properly switch between (until very recently). USA considers this a 'success', because we 'deployed early'.

 

Europe properly sees that as a failure. The major goal of their regulation is to prevent exactly what happened in the USA. LTE intentionally being separately and incompatibly banded is a *bad thing* for *everyone*. It's anti-competitive to lock phones down like that, and it encourages isolated networks that force extra costs onto consumers.

 

A properly run FCC should have pushed operators to do the right thing here. A properly run FCC should banded all 700mhz spectrum so that LTE devices on those bands were entirely cross compatible between any network.

 

- - - -

 

I think it's very much the 'pot calling the kettle black' when we point at Europe's "slow" adoption of LTE. In the United States, we still sell proprietary, archaic, legacy 1x/EVDO devices as modern equipment, and are still *deploying* new EVDO and 1x CDMA service on sites. That's the definition of backwards, even Canada is beating us in this regard, having properly switched over 95%+ of their population to GSM/HSPA+ by now.

 

We can puff up our LTE deployment stats all we want -- but by international standards, USA cell service is still a giant mess of crazy and incompatible bands and services. I would greatly prefer we adopt a more European approach, if it meant providers were forced into modern GSM services and compatible band planning.

What people don't understand is that telecommunications is all about scale and interoperability. That's how the original phone system managed to grow throughout the world. Cry about it all you want, but there's a reason that universal service typically demands universal technology and compatibility.

Despite the fact that our networks are well-built out, it hasn't been long enough that the dust has settled yet. We still have an opportunity to push interoperability, better spectrum management, and better technology management. A lot of us tend to forget that these companies don't own the spectrum (even though they make it sound like they do). We do. 

I consider many telecom companies out there to be very hypocritical. Even at the CCA, where the smaller carriers bemoan the lack of interoperability and stuff like that, they talk out of both sides of their mouths. They want the big carriers to not have that ability, but they want it to screw over their customers. They've even said that they don't want to offer that to their customers.

A.J. often accuses me of having a Eurasian view to telecom. And perhaps he's right. But the Eurasian view is that common technology, common frequencies, and common networks are incalculably beneficial to the public. CEPT has done a great job of that in Europe, and even though the spread of that beyond Europe is tinged by some darker undertones in Africa and Asia, I think it still worked out fairly well.

Contrary to popular belief, I liked using CDMA2000 when it first arrived. It was superior to FD-TDMA systems like D-AMPS and GSM. But, I switched to UMTS as soon as it arrived, because I felt that it sufficiently incorporated the advantages of CDMA technology with some advancement and interesting capabilities of its own. And of course, it was a global technology being adopted by all regions. I've never made it a secret that I prefer UMTS over CDMA2000. I personally like that operators in Canada, Latin America, and most of Asia have switched from CDMA2000 to UMTS as they've realized the socioeconomical weakness of the CDMA2000 platform. For a company like Sprint, who is deploying multi-mode gear, it's very easy to add UMTS to it and transition over time to UMTS/LTE. T-Mobile is doing the same thing in MetroPCS CDMA/LTE markets, and it's been going along very swimmingly. I think that operators in the US should move to a common UMTS/LTE platform, especially since our spectrum for LTE is so screwed up that we've practically broke the key benefit of everyone using LTE.

I've also never made it a secret that I think the FCC needs a swift kick in the pants by the public to force them to fix our spectrum problem. We don't have a spectrum shortage problem. We have a spectrum mismanagement problem. This is because Congress keeps forcing the FCC to auction spectrum when it isn't ready, and the FCC keeps taking shortcuts to auction spectrum, leaving it fragmented and useless. Thankfully, the FCC doesn't make totally insane decisions (like making 600MHz spectrum TDD). However, the FCC's work with 700MHz, S band, PCS H, L band, etc. prove to me that we, as the public, need to force the FCC to fix this -- once the government is back online.

Whoever said that Legere just hates AT&T that much, probably nailed it. Paying a little more for roaming would help out T-Mobile immeasurably but he doesn't want to do anything to help his customers, because he'd have to admit that AT&T awoke from their network slumber and started taking their network building seriously. I know the network end takes time since T-Mobile has to do rebuilds of almost all of their rural sites. A little relief in the meantime would be nice.

And where did you get the idea that Legere hates AT&T? For goodness' sake, he learned the ropes of telecom while working there, alongside Dan Hesse and other stalwarts in telecom today!

I doubt he hates AT&T. However, it's a sound strategy to pick a fight like he's doing. The "us vs them" mentality is easy to fall into, and Legere is making it "us" to be T-Mobile, and "them" to be AT&T. Invoking that "underdog" and "challenger" imagery with T-Mobile as the "good guy" and the other three as the "bad guys" helps rally the troops and it is helping to turn around the company.

Sprint is no stranger to this, but it hasn't been executing well to back up its marketing. So it has somewhat fallen flat.

I'm fairly certain this is not true, AT&T roaming is never so expensive that T-Mobile could not afford it.

 

I don't have hard numbers to back it up offhand, but numerous sources (including T-Mobile's own press release) state that the roaming agreement rates were "favorable to T-Mobile"

 

T-Mobile may choose to not pay for roaming in certain areas, for whatever reason they'd like. (AT&T 'charging to much' likely means too many T-Mobile customers would rack up AT&T roaming charges that T-Mobile chooses not to pay, not that AT&T network access is so expensive on a given tower that T-Mobile can't afford to pay them).

 

Besides, it's not like they could't charge for domestic roaming. Lots of subscriber would pay an extra $10/month just to have 50mb to 100mb of nationwide AT&T data roaming. (Just as Sprint charges extra for Verizon data roaming).

 

http://gigaom.com/2011/12/20/t-mobiles-consolation-prize-a-whole-lot-of-airwaves/

http://www.engadget.com/2011/12/20/t-mobile-usa-and-atandts-seven-year-umts-roaming-agreement-gets-d/

Favorable is something as simple as being able to pick and choose what areas would have roaming enabled, and flexibly calculate the fees incurred based on that.

I don't quite remember the specifics, but I thought part of the failed AT&T deal was a flat roaming rate? 

No. No one mentioned a flat roaming rate. However, T-Mobile has the option of enabling AT&T voice and data UMTS roaming anywhere it can afford to. AT&T gets to rate-limit T-Mobile to CSD speed, though. 

Some in-market areas are allowed. Lots of out-of-market areas are blocked...

 

There seems to be zero rhyme or reason behind it.

 

It's like they threw darts at a map one day and said "in these areas, and only these areas, we will allow AT&T roaming". And it's stuck that way ever since.

 

It's ironic, because it would save them a lot on the coverage issue. Sprint allowing roaming on Verizon pretty much solved the 'coverage' issue, because anytime someone worried about coverage, Sprint could just say 'no worries, we roam on Verizon for 'free', your never without coverage'. With the flip of a switch, T-Mobile could do the same with AT&T.

 

I think John Legere is preventing it -- It would require him to admit they have better coverage, and he's physically incapable of saying anything about AT&T that could ever possibly be interpreted as a positive.

If T-Mobile could afford to enable roaming across the entire country, it would. But AT&T does not offer roaming on a national basis for domestic roaming. Roaming rates differ based on location. In some areas, AT&T simply charges too much to enable any roaming. In others, AT&T charges a more reasonable rate.

Sprint's situation with Verizon is unusual, and has not been repeated by anyone else since the original Alltel disappeared. 

So T-Mobile users now have unlimited data roaming almost anywhere in the world... except in America, where you're stuck with a 50mb roaming limit with no option to pay for more...

 

...But even worse- you are extremely limited on where you can actually roam, and most of the time you will only have "emergency calls only"

Unfortunately, domestic roaming agreements don't work the same way that international ones do. The closest analogue would be India, where T-Mobile has signed agreements with multiple operators to get pan-India coverage. And even then, the operators don't give T-Mobile any indication whether they truly offer that.

Tmo is pretty much just software upgrades for LTE where as sprint is completely new hardware and backhaul, pretty much building from the ground up right?

Not entirely. Some places are like that, but others are a modernization effort like Sprint's. 

I didn't mean to come across like you weren't aware of it. I just wanted to emphasize to anyone reading the thread that making the business case for them helps. 

Oh, okay.

No harm done. 

Heartfelt plus making a good case can sometimes go far.

I'm well aware of that. I did it myself to get Starkville, MS upgraded by T-Mobile last year. I'm grateful to the T-Mobile executive and engineering teams for listening to me and doing that.

We often forget that businesses are run by people, and telecom is a people-oriented business (despite how many feel about it right now).